Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feat(interactive): add use-mmap-vector and build-csr-in-mem option for bulk loading #3957

Merged
merged 27 commits into from
Jun 20, 2024

Conversation

liulx20
Copy link
Collaborator

@liulx20 liulx20 commented Jun 19, 2024

When loading the edges of a graph, there are two stages involved:

  • The first stage involves reading the edges into a temporary vector and acquiring information on the degrees of the vertices;
  • Then constructs the CSR using the degree information.

During the first stage, the edges are stored in the form of triplets, which can lead to a certain amount of memory expansion, so the use-mmap-vectoroption is provided, mmap_vector utilizes mmap to map files, supporting runtime swapping memory to disk.

Constructing the CSR involves random reads and writes, we offer the build-csr-in-mem option, which allows CSR to be built in-memory to avoid extensive disk random read and write operations.

@liulx20 liulx20 merged commit 9adeefc into alibaba:main Jun 20, 2024
36 of 37 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants